nick black

also, if you get cancelled in a condition variable wait, don't you reaquire the lock? i think you do. yeah: >CANCELLATION > pthread_cond_wait and pthread_cond_timedwait are cancelation points. If a...

yep, looks like that got it, w00t! not sure if this is the same thing responsible for the macos ci hang but i suspect it is.

looks like we just hit this again at https://github.com/dankamongmen/notcurses/actions/runs/13054111564 arrrrrr

ok the macos issue does not necessarily look input related. i've got a (huge) logfile. ``` 2025-01-30T14:09:50.1344880Z [100%] Linking CXX executable notcurses-tester 2025-01-30T14:09:50.2092680Z [100%] Built target notcurses-tester 2025-01-30T14:09:50.2245860Z ##[group]Run cd...

so can we not run with `output-on-err` or whatever it is here? also, we get some compiler warnings on the macos build, it seems: ```c Z /Users/runner/work/notcurses/notcurses/src/lib/unixsig.c:7:34: warning: macro 'ATOMIC_VAR_INIT'...

```c 2025-01-30T14:09:38.8099230Z /Users/runner/work/notcurses/notcurses/src/tests/metric.cpp:95:5: warning: 'sprintf' is deprecated: This function is provided for compatibility reasons only. Due to security concerns inherent in the design of sprintf(3), it is highly recommended that...

thank you for the report! i'll look into this shortly.

do you know whether this is glibc or musl? i'm assuming the former

there's no way you can get me a valgrind run on this, is there?

so i'm not familiar with github codespaces. your `TERM` is `xterm-blah`; is there a an actual XTerm involved here, or is this some browser deal? we oughtn't be dropping core...